Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
29.8FC F1
13
May 27, 2026
29.8F1 (Factual Correctness)
13
May 27, 2026
94.9Accuracy
13
May 26, 2026
51.16Initial Score
13
May 20, 2026
71.2BLEU
13
May 12, 2026
95.9Accuracy
13
May 8, 2026
62.545Accuracy
13
Apr 28, 2026
46.3Exact Match (EM)
13
Apr 21, 2026
65.8EM
13
Apr 21, 2026
47.6EM
13
Apr 21, 2026
83.8Accuracy
13
Apr 21, 2026
52.53Accuracy
13
Apr 16, 2026
65.44Image F1
13
Apr 16, 2026
27.14ANLS
13
May 20, 2026
0.25ANLS
13
May 20, 2026
21.43ANLS
13
May 20, 2026
73.56Normalized Accuracy
13
Jun 8, 2026
71.1EM
13
Feb 26, 2026
82.9Accuracy
13
Feb 26, 2026
89.2MMLU Accuracy
13
Jun 18, 2026
47.4SAcc
13
Feb 26, 2026
63.9Accuracy
13
Feb 26, 2026
75.5MMLU Accuracy
12
Jul 8, 2026
22.4F1 Score
12
Jul 3, 2026
4.94F1 Score
12
Jul 3, 2026